Suggested readings for Week 5

If you’re a regular reader of this blog, you probably have some interest in language–language in the abstract, or “language” in French, as opposed to (or in addition to) any particular langage, or “langue” in French. I’m reblogging here the reading list for a week of a course on language processing that I’m teaching at the moment. The theme is data in language processing: what you (might) mean when you talk about “data” with respect to language; what kinds of data there are; where that data comes from; and how to make it if you can’t find the kind of data that you need.

I’m posting this particular reading list because I often suspect that many people who know that I’m a linguist imagine that I spend my days sitting around discussing how funny irregular verbs are, or how cool it is that French has three verbs that mean “go back,” or whatever. What you’ll find on this list has very little to do with coolness or lack thereof, and a lot to do with data formats, data set sizes, statistics, and a bit on ethics. Personally, I find this stuff fascinating–but, it’s often worth getting a glimpse at what we call in my field “the sausage-making process.” Enjoy! (Or go watch the latest episode of “The Walking Dead”–it’s pretty good.)

Natural Language Processing

Here are some suggested readings for Week 5.  Remember that I do not distribute my lecture notes.  Note also that you are responsible for all of the material on which I lecture.  These readings are not required, but they are intended to cover everything that I talk about in our lectures (modulo the caution in the preceding sentence).  All of them are available for free on line except for the books (although the Good and Hardin book is available for free, as well).  All of them should be available in an academic library.  Feel free to contact me if you have trouble finding a copy of either.

View original post 360 more words

The zombie apocalypse and education in the computational sciences

How to respect both logical positivism and the zombie apocalypse while educating computer scientists.

Screenshot 2017-03-10 04.25.01, a web site that supports research on what linguists call the “heads” of groupes nominaux (“noun phrases,” in English).

In my professional life, one of my pet peeves is scientific discussions that involve the verb to believe.  For example:

  • …we believe that [joint circumscription] will be important in some AI applications.  (John McCarthy, Circumscription–A form of non-monotonic reasoning, publication date unclear) 
  • We believe ontologies are key requirements for building context-aware systems… (H. Chen, T. Finin, and A. Joshi, An ontology for context-aware pervasive computing environments, 2003)
  • We believe enzyme-loaded erythrocytes may have therapeutic possibilities for several diseases.  (Ihler et al. 1973, Enzyme loading of erythrocytes, which I should note has been cited over 300 times nonetheless)

I have actually been–on multiple occasions–cautioned against using formulations like Je pense que… (“I think that…”) in some professional situations in France, as it’s considered a sign of having a position that you’re not actually confident that you can defend.  (Native speakers, can you comment on this?)

I’m not shy about bringing up my problems with the verb to believe in any discussion in which I find myself that claims to be scientific, be those lab meetings or reviews of papers/grants/whatever.  I would not label myself as a logical positivist, but I try to always keep in mind the potential logical positivist position–it’s not a bad foundation for a philosophy of science.  (See, I didn’t say I think that it’s not a bad foundation for a philosophy of science–I flat-out asserted it.  In academic writing, I would follow that assertion with a few credible citations.)

Follow these links for more information on the zombie apocalypse and…

In light of that tendency of mine towards the empirical and the epistemological, students are often surprised to learn of my concerns regarding the upcoming zombie apocalypse.  Clearly, zombies are something about which I have no empirical data, and one would have to classify the upcoming zombie apocalypse as something about which I have beliefs, but not knowledge, and therefore outside of the realm of something that I would talk about in my professional life.  So, yes: students are surprised when I bring it up.  (As far as I can tell, my French colleagues just think I’m crazy, or chalk it up to some quirk of the Anglo-Saxon psyche, or something.  I actually have no clue what my American colleagues think.)

Here’s the thing: the zombie apocalypse is an engaging point of entry into the problem of making robust systems.  In the context of computer programming, you could think of “robustness” as the ability of a program do deal with the unexpected–making speech recognition systems that will work in a crowded restaurant (impossible 20 years ago, not unusual today), or building sentence analyzers that won’t reformat your hard drive if someone passes them a sentence in Uzbek. In particular, the upcoming zombie apocalypse is an engaging entry point to the problem of how to think about the problem of making robust systems.  The issue is that a major contributor to robustness is planning for unanticipated inputs (I had English in mind when designing my sentence analyzer, and then someone gave it a sentence in Uzbek) or operating conditions (I never thought about someone trying to use my speech recognition system with a lot of noise in the background).  Seulement voilà–the thing isit’s the nature of unanticipatedness that we have trouble coming up with the unanticipated.  Even more fundamentally a problem: we often have trouble getting into the mindset of taking seriously the very idea that unanticipated inputs or operating conditions are even plausible.  In fact, they are; but, how to get students to think about something that is, a priori, difficult to conceptualize?  Posing the question as how will your approach work when the zombie apocalypse comes? typically leads to a laugh–and seems to give one a way to think seriously about what kinds of things might happen that you haven’t actually thought about yet.  To think seriously about things that it’s difficult to think about by means of thinking non-seriously about things that don’t exist, you might say.  You might say that–if you haven’t really thought about the upcoming zombie apocalypse.

English notes

pet peeve: something that annoys a specific person a lot.  To call something a pet peeve, it should be rather specific to that person, especially with respect to the extent that it bothers them or with respect to the extent that they are sensitive to it.  For example, traffic jams wouldn’t really be anybody’s pet peeve–everybody is annoyed by traffic jams.  However, traffic jams caused by trash trucks doing their collections during rush hour could be someone’s pet peeve-say, if they happen to actually notice them more than most people would, in a situation where most people don’t particularly care whether or not a traffic jam was caused by a trash truck doing its collections during rush hour–they are equally annoyed by all traffic jams.  How it was used in the post: In my professional life, one of my pet peeves is scientific discussions that involve the verb “to believe.”  

French notes

la robustesse: robustness.  You can use this in a lot more ways in French than in English.  For example:

  • Hardiness would probably be the English-language equivalent here, where we’re talking about plants and their illnesses: Différentes maladies peuvent entraîner un flétrissement des tubercules qui se traduit, à son tour, par une perte de robustesse des plants.  (Source: Sketch Engine web site)
  • Toughness would probably be the equivalent here, where what’s being discussed is fabric: Ce tissu se distingue par sa robustesse, sa longévité et son confort.  (Source: click here)

Да, да: How to irritate a linguist, Part 3

There are soooo many ways to annoy a linguist–here’s a good one.

bulgarian flag bg_971
Favorite thing that I ever heard anyone say in Bulgarian: “Ani says that our lipstick should be as red as the bottom of the flag.”  Picture source:

My airport shuttle pulled into a truckstop in Bulgaria, someplace between Sofia and Hissarya, so that we could all get out and stretch our legs.  I paid the nice little old bathroom attendant my 50 stotinki and walked to the stalls, feeling proud of myself because I knew how to recognize the men’s room and the ladies’ room in Bulgarian Cyrillic.  My hand froze as I reached for the door when I heard the nice little old lady scream другата, другата!!!–I spoke just enough Bulgarian to know that that means the other one, the other one!!!  You know what they say: pride comes before a fall.

For your amusement, here is an article with many correct observations, and nothing but incorrect conclusions.  The article is about the English language.  The incorrect conclusions start with the first sentence, where one is stated quite clearly, and then they continue through the body of the piece, becoming increasingly more academic/less clear.  (En clair: “in plain language, in plain English; not scrambled (e.g. a TV show).”)  For my amusement, here are some comments while I wait for the coffee to finish.  Quotes are in italics.

First sentence: English is a difficult language.  What would it even mean for a language to be “difficult?”  That children can’t learn to speak it natively?  No such language exists.  That you had trouble with it in high school?  That’s not exactly a convincing form of evidence.  That you think that people with differently-colored skin/from different parts of the country/from different social classes than you don’t speak it correctly?  Fuck you.

Second sentence: It’s irregular (teachers taught, preachers praught?), single words take on multiple meanings (‘set’ has 464 definitions in the Oxford English Dictionary) and its pronunciation is fiendish (cough, though, bough, through, tough, plough).  I love the “teachers taught, preachers praught” thing–super cute–but have no idea why one would think that irregularity was any more difficult than anything else.  From a hearer’s perspective, the discriminative power of irregulars is much higher.  To think about it from a computational perspective: irregulars don’t require any processing–you just consult the equivalent of a mental “dictionary.”  In contrasts, regulars require that you chop things up (e.g. if it were teached, you’d have to separate teach and -ed) and then figure out what they were (the -ed of kissed is not pronounced the same as the -ed of hugged), and then look up what they meant.  You can argue that the claim that I’m making here about increased discriminative power for the hearer means increased memory load for the speaker, and you’d be right–which is why linguists don’t bother talking about this kind of thing very much.  You quickly come back to the question that I raised about the first sentence: what would it even mean to be a “difficult language?”

The author continues:

This isn’t only about our famous English sarcasm. It’s a sign that the words themselves are not as reflective of our thought as, perhaps, they should be. When I say, ‘I beg your pardon’, I can mean ‘I apologise’, ‘I didn’t hear you’, or most probably ‘I’m absolutely fuming at what you said’. No wonder English is famed for being such a tricky language to learn, if what we’re trying to get across is based really on our tone of voice more than anything else.

Does the author really believe that this is something that is unique to English?  As far as I can tell: yes, he does.  But: it isn’t.  Here’s an example from Bulgarian:

Да, да.

What those words are: yes, yes.  What that means: No.  You can tell the difference from the intonation.  Is that kind of thing interesting?  Absolutely.  Is it in any way unique to English?  No.

…single words take on multiple meanings…  From my perspective, that is a confusion about what a “word” is.  If a word is a relationship between a sound and a meaning, then the thing to say here would be that there are multiple words in the language, some of which sound the same as other words.  If that’s the case: big deal.  Do you know of any languages where that’s not the case?  I don’t.

…and that, my friends, is yet another way to irritate a linguist: say dumb shit about how special Language X is because you don’t know anything about any other language, and therefore don’t actually have anything to compare it to.  (The author does mention Mandarin, but what he says about it is too stupid uninformed for me to respond to before I’ve had another cup of coffee.  Back reviewing grant proposals–computational linguistics isn’t all beer and pétanque

How defending Trump is like defending domestic abuse

From a linguistic perspective, defenses of Trump have a lot in common with domestic abuse. Here’s how that works.


Data point: back in the United States, our new President has been gleefully violating our Constitution, or at least trying to, to the very best of his ability.  The heart of American political philosophy is expressed in the First Amendment to the Constitution:

Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.

What our new President’s been up to:

  • Banning people from entering the United States based on their religion.  Quote from December 7th, 2015: Donald J. Trump is calling for a total and complete shutdown of Muslims entering the United States.  (Yes, he is enough of an asshole to refer to himself in the third person.)
  • Attacking the press.  Tweet from February 17th, 1:48 PMThe FAKE NEWS media (failing , , , , ) is not my enemy, it is the enemy of the American People!

It interests me that many of his supporters defend his un-American actions based on the argument that he’s “just” doing what he said he would do.  There seems to be some implicit claim that if you say that you’re going to do it, then it’s OK to do it.  Some examples:

  • “He was simply doing what he said he was going to do in the campaign,” Paul Hess told the Times.  (source)
  • President Trump is, after all, just doing what he said he would do. And, in a representative democracy, that’s something to be respected.  John E. Stafford, letter to the New York Times
  • Amid all the wailing and gnashing of teeth, President Trump reminded us that he was just doing what he said he would do.  “We have really done a great job. We’re actually taking people that are criminals, very, very hardened criminals in some cases, with a tremendous track record of abuse and problems, and we’re getting them out,” Trump said. “And that’s what I said I would do. I’m just doing what I said I would do.”  Sean Hannity (source) (includes video of Trump saying exactly that)

As far as I know, this is not the case–either from a legal perspective, or from an ethical perspective.  If I say to you If you drink my Coke again, I will punch you in the face, I’m going to arrested if I do, in fact, punch you in the face–having said in advance that I was going to do it does not make it legal.  It does not make it ethical, either.

“I warned her I would kill her if she went with other boys,” he added. He said that Sunday afternoon she went to a show with another boy and that “she broke her promise at other times.” “I kept my promises and she broke hers. I loved her very much,” he added.  –Source:

The whole phenomenon reminds me of the stereotype of domestic abusers: This is your fault–I told you I would hit you if you talked to him again.  I told you I would whip you if you didn’t come straight from school.  I told you I would kill the kids if you tried to leave me. Do domestic abusers actually do that kind of thing?  Read the quotes.

Now, there’s an interesting little linguistic thing going on in the quotes from the Trump defenders.  Let’s look at the quotes again:

  • “He was simply doing what he said he was going to do in the campaign,” Paul Hess told the Times.  (source)
  • President Trump is, after all, just doing what he said he would do. And, in a representative democracy, that’s something to be respected.  John E. Stafford, letter to the New York Times
  • Amid all the wailing and gnashing of teeth, President Trump reminded us that he was just doing what he said he would do.  “We have really done a great job. We’re actually taking people that are criminals, very, very hardened criminals in some cases, with a tremendous track record of abuse and problems, and we’re getting them out,” Trump said. “And that’s what I said I would do. I’m just doing what I said I would do.”  Sean Hannity (source) (includes video of Trump saying exactly that)

Note the words simply and just.  They have a very specific function here.  You can think of it as justification through minimization: their goal is to communicate the idea that what follows is not a bad thing, specifically by minimizing it relative to things that would admittedly be bad.  It’s quite complex, because it starts with a concession–with an implicit agreement that if you had done what the other person states that you did, then it would have been bad.  But, that concession is then followed with an argument that what you did was not, in fact, that, and since it wasn’t that, then you are meant to accept that it was not bad. 

The English words just and simply have a lot of meanings.  This particular meaning gets used in a couple very particular ways.  I’ll give you the more complicated one first, because it’s actually easier to see how they function in the more complicated case:

  • Yes, I did sorta take some of your sandwich, but I really just tasted it.
  • Mom, I didn’t hit him–I just touched him hard.
  • Don’t get mad at what I said–it was just a joke.
  • I’m not being mean–I’m simply stating a fact.  You are fat, old, and bald.

The structure of all of them works something like this: there’s this thing that you think that I did, and if I had done it, sure, maybe that would have been bad.  But, I didn’t–I did some lesser thing, and since it’s not the bad thing that you mentioned, it’s not bad.  So:

  • Yes, I did sorta take some of your sandwich, but I really just tasted it.  The structure: if I had really taken some of your sandwich, then that would have been bad–but, I didn’t.  I did a lesser thing: I tasted it.  Since that’s not the bad thing, then it’s not bad.
  • Mom, I didn’t hit him–I just touched him hard.  The structure: if I had hit him, then sure–that would have been bad.  But, I didn’t hit him–I did something less than that, and since it’s not that bad thing, then it’s not bad.

This is a fallacious argument.  Suppose that there is some bad thing–let’s call it X.  The fact that something is not X does not mean that it is not bad.  The fact that something is “less” than X does not mean that it’s not bad, either.  But, that’s exactly the implication behind the whole “he’s just doing what he said he would do” attempt at a justification.  In fact, “just” is being used here without the concession–it’s pure minimization.  It’s adding to the assertion he’s doing what he said he would do an adverb that is meant to convey that the doing is something less than something else–specifically, less than bad.  

At some point, the current insanity is going to end–America always rights herself, eventually.  How?  Who knows?  Maybe Trump will throw one of his little hissy fits and resign.  Maybe he’ll nuke somebody, and somebody will nuke him, and the world as we know it will end.  These days, it’s tough to be surprised.  One thing that I am, however, sure of: history is not going to look kindly on this period, and it’s not going to look kindly on the people who supported Trump.  Are they all going to go to jail?  Of course not.  Are their grandchildren going to be ashamed of them?  Probably.  You have a choice to make here: collaborate, talk back, or just keep your head low.  There’s only one of those that you won’t be ashamed to tell your grandkids about.

Semiotic analysis of a Heinz Ketchup advertisement

We’ve talked before about the study of semantics (meaning in language) versus semiotics (how things have meanings). We’ve also been talking recently about specific ways that language can be used to manipulate opinion. The article that I’m “reblogging” here is a nice example of using not so much language and its semantics, but rather the broader field of semiotics, to manipulate opinion, courtesy of the Advertising and Society blog.

This ad from 1900 is a good example of a common strategy for getting people to buy a commodity from you: the ad tells you nothing about Coca-Cola, but plays on your desire to want to be like the lady in the picture–young, wealthy (note the elaborate clothing), and loved (note the flowers). Picture source: public domain,

The post is about an advertisement for ketchup. Ketchup is a good example of what’s called a commodity–a product with the property that it generally doesn’t really matter who you buy it from, because it’s mostly all the same. Other classic examples of commodities are wheat, sugar, toothpaste, and razors. If you are, as they say, in “commodities hell”–trying to out-compete other people when all of your products are pretty much the same–then you have to convince people to buy the product from you on the basis of something other than the product itself. Typically, that involves painting a picture in which your product is associated with something that your customers value other than the product itself–family, love, or in this case, health.  Portwood-Stacer’s post is an extended analysis of a very simple-looking advertisement that makes use of semiotics in a pretty sophisticated way, and in particular the interplay between symbols (Porter-Stacer refers to them using a technical term, sign) and the way that symbols can take meaning from their societal context.  Enjoy, and thanks, Dr. Portwood-Stacer!

Advertising & Society

Semiotic analyses of advertisements reveal cultural norms and values associated with a particular society or group of people. In fact, in order for people to decode signs they must do it within their own sign system (dependent on language, historical context, and culture). Social Communication of Advertising, writes, “Semiotics highlights the way that we ourselves take part in the creation of meaning in messages, suggesting that we are not mere bystanders in the advertising process, but participants in creating a code that unites the designer and reader” (Leiss, Kline, Jhally, Botterill 164). Advertisers depend on these signs in order to communicate a point quickly and effectively to consumers.


In the above print advertisement for Heinz ketchup, the signifiers include a vivid red backdrop, a classic bottle of Heinz ketchup horizontally sliced with a tomato on top, and white text reading “No one grows ketchup like Heinz.” While a seemingly…

View original post 408 more words

How to not get a second date with a non-linguist

I always love being in France–but, sometimes, I REALLY love being in France.

The first thing you learn in American linguistics graduate schools is that you can make sure that there will never be a second date by commenting on some aspect of your companion’s speech.  Although America has no official language and nothing remotely like the notion of an Académie-Française-sanctioned standard form of the language, we are nonetheless super-sensitive about having the way that we speak brought up in a conversation.  Comment on the way that your date speaks, and it’s all over.  In France, the situation is very different–anyone will talk about how anybody else speaks, anywhere, any time.  I love that.

Sunday is market day in my little neighborhood in Paris.  Vendors set up their booths under the metro tracks down the block (I live by one of the few “aérienne” (elevated) lines).  Most things are pretty local–in France, meat and produce is usually sold with its area of origin marked, and the majority of foodstuffs for sale at the market come from no further away than Spain.  (For my geography-challenged American concitoyens: that’s right next door.)

I have my little routine.  The first place where I stop is the aligot booth, because if they were to sell out of that potato-butter-and-cheese equivalent of crack cocaine before I got there, my week would be ruined.  On the last leg of my trek, I stop by the choucroute stand.

Choucroute is an Alsatian specialty consisting of sauerkraut, an occasional carrot or potato, and any of a wide variety of smoked and/or cured meats.  Which raises a question: which meat do you want?  My habitual choice: all of them.

Picture source:

When I got to the front of the line for the choucroute, the elderly gentleman next to me was having a detailed discussion with one of the ladies working the booth about the ham on offer, and exactly how close to the bone it had been sliced.  The lady had set the pig leg on the counter, and was indicating various and sundry parts of the unfortunate animal’s anatomy with her knife.  (How close to the bone you’ve been sliced turns out to have implications for how deeply the meat has been cooked, and therefore both the smell (apparently worse the closer you get to the bone) and the taste (apparently better the closer you get to the bone).)  I looked at the variety of meats resting atop the bed of fermented cabbage and decided, as I usually do, that I wanted a bit of everything.  (If it’s in italics, it happened in French.)  May I have a mix of meats, please?  A huge smile from the vendor: oh, what a beautiful French word!  Did you hear what he just said?  …she asked the gentleman examining the ham.  He grunted and went back to discussing bone-closeness.  Shit, I thought to myself–what did I just say??  

Where are you from?  America, really?  Seriously, did you hear?  He said “déclinaison de viandes.”  This time the elderly gentleman didn’t even bother to grunt–nothing was going to distract him from his deepening relationship with that ham.  What should I have said? …I asked.  A “mélange,” I think…or an “assortiment.”  But, don’t change–that’s delightful.  

Lest you think that I’m bragging: this wasn’t the last time that I amused the nice choucroute lady yesterday morning.  In particular, when she asked me if I wanted some alaille, I was baffled.  She was happy to explain to me that this was saucisson à l’ailgarlic sausage.  D’oh!  On the down side, I still sound like a complete idiot when I try to speak French.  On the plus side, I gave the nice choucroute lady a few good laughs, and that has to count as A Good Thing.  I always love being in France–but, sometimes, I LOVE BEING IN FRANCE.  Seriously.

English notes

atop: a preposition meaning on top of.  This is a word that you might use in writing, but would rarely, if ever, use in the spoken language.  How it was used in the post: I looked at the variety of meats resting atop the bed of fermented cabbage and decided, as I usually do, that I wanted a bit of everything.

trek: a long journey, usually done specifically by walking, and usually difficult.  How it was used in the post: On the last leg of my trek, I stop by the choucroute stand.  In this case it conveys the idea that my journey through the market is long, and that I’m walking, but in this context, it’s not meant to suggest difficulty.

French notes

la déclinaison: according to WordReference, a range or variation; I saw it used in this way on the ardoise (“slate”–the little blackboard, often an actual piece of slate, on which the specials of the day are posted in restaurants) of the cafe downstairs from my apartment, advertising a déclinaison de tomates–an assortment of tomatoes.  Also according to WordReference, a declension, in the sense of a set of related words (sausage/sausages/sausage’s).  I have loved this word from the moment that I learnt it–apparently the choucroute lady thinks it’s pretty cool, too.

Questions with only one right answer

You’re in country X.  Let’s say that the local language is called Xish.  Here are the only correct answers to the following questions:

Q: So, what do you think about Xish?  A: It’s beautiful.
Q: Xish is really easy to speak, isn’t it? A: No.
Q: Do you think that Xish is hard?  A: Yes.
 Q: What’s more difficult–English, or Xish?  A: Xish.
 Comment: You speak Xish wonderfully!  Response: Oh, no, I speak Xish terribly.

In some technical sense, your answer to all of these will have been been false, except for the one about speaking Xish poorly.  “Difficulty” is not a meaningful word when applied to languages.  Neither is “beauty” in a technical sense, although I won’t belabor that one.

It occurred to me as I wrote this that the picture that I’ve painted here could be interpreted as suggesting that people who speak any language other than the one that you speak are easily fooled. In fact, that’s not the case at all.  This is about shared human culture–as far as I know, most people in most places love to talk about their language with foreigners, and how hard that language is will pretty much always be a good conversational tack to take.  (Obviously, I haven’t been everywhere or talked to everyone, but I’ve probably done this little exercise in somewhere in the neighborhood of 20 countries by now.)  In fact, in a lot of places, the Your Xish is great! thing is a sophisticated opportunity to let you show your grasp of the culture (or not)–in many cultures, accepting a compliment is quite gauche, and the only proper response is self-deprecation.  Respond with “oh, thank you–I’ve really been working on it!”…and you’ve just shown yourself to still be clueless.  Respond appropriately and you’ve just shown your grasp of, and respect for, the culture.

Ironically, I can’t quite figure out whether or not that’s the case in France–in general, this is not a country where self-deprecation is valued.  It’s a real problem for Americans, since self-deprecation is more or less our default attitude any time that we meet someone new, and often for much, much longer than that.  You could think of this whole isn’t-my-language-hard thing as an instance of not “exoticizing the Other,” as we academics like to say, but rather, of exoticizing oneself–of supporting a sort of exceptionalism for one’s own language, in the sense that we talk about “American exceptionalism” (the idea that America is just plain better than the rest of the world and has something to offer it–I certainly agree with the second part of that) and “French exceptionalism” (the idea that France is just plain better than the rest of the world and has something to offer it–I certainly agree with the second part of that, too).

English notes

Picture source:

gauche: lacking social experience or grace; also :  not tactful :  crude  (taken directly from Merriam-Webster).  I think that the best French equivalent might be maladroit, but couldn’t swear to it.  How it was used in the post: In many cultures, accepting a compliment is quite gauche, and the only proper response is self-deprecation.  Some examples from the Open American National Corpus, a collection of 15 million words of American English, collected and annotated by my colleague Nancy Ide, that you can download to do with as you please.  I used the Sketch Engine web site to search it.

  • You are correct that you cannot come right out and say, “It is gauche to come over and serenade me with your potato chips , so please go away.”
  • Gauche, gauche, gauche, and tacky.  (I love this one even more than the previous one.) 
  • Your take on his behavior was correct: It was gauche. Prudie does have one slight bit of curiosity about the faux pas.

to take a tack: to go in a particular direction, metaphorically speaking.  It comes from nautical language, where the verb to tack means to change the direction of a ship by turning the bow into the wind.  Confusingly, it can mean something like tactic, but it is not related to that word at all.  How it was used in the post: Most people in most places love to talk about their language with foreigners, and how hard that language is will pretty much always be a good conversational tack to take.  Some examples from the Open American National Corpus (see above):

  • Britain’s Independent took a similar tack, observing, “The situation is far from precisely parallel, but it is still a chastening thought that the Kosovo Liberation Army is, under conditions of vastly greater duress, handing in its guns at a rather faster rate than the Provisional IRA seems able to arrange  “
  • But rather than pursue that obscure tack any further (place names such as Washington are surely both proper nouns and eponyms) , let us see if the proper categories of words really end there as grammar books tend to suggest .  (Different verb–pursue, rather than take–but, same meaning)
  • Having apparently grown tired of obsessing over just how skeletal the Ally McBeal Über-waif has become, the tabs take a different tack: They bare their fangs and become positively McCarthyesque in their zeal to rat out celebs who’ve become the least bit unsvelte.
  • I think it’s one of the tacks Gerald Posner took in his book JFK book, Case Closed.

French notes

gauche: according to, this adjective can mean awkward, clumsy, or gauche, but with this sense (meaning) it is soutenu.  

le langage soutenu or le registre soutenu: according to the French-language Wikipedia, this is especially a written form of the language, used in official letters and literary texts.